COMMUNITY ECOLOGY 14(1): 67-76, 2013 1585-8553/$20.00 © Akadémiai Kiadó, Budapest DOI: 10.1556/ComEc.14.2013.1.8 Arbuscular mycorrhizal colonisation of roots of grass species differing in invasiveness G. Endresz1,2,4, I. Somodi3 and T. Kalapos1 Department of Plant Systematics, Ecology and Theoretical Biology, Institute of Biology, Eötvös Loránd University, Pázmány Péter sétány 1C, 1117 Budapest, Hungary Present address: Szent László Secondary School, Kõrösi Csoma S. út 28-34., 1102 Budapest, Hungary ! MTA Centre for Ecological Research, Institute of Ecology and Botany, Alkotmány u. 2-4., 2163 Vácrátót, Hungary " Corresponding author. Email: [email protected] Keywords: Arbuscular mycorrhizal fungi, Calamagrostis epigejos, Cynodon dactylon, Grasses, Invasive plants, Semiarid temperate grassland. Abstract: Recent research indicates that the soil microbial community, particularly arbuscular mycorrhizal fungi (AMF), can influence plant invasion in several ways. We tested if 1) invasive species are colonised by AMF to a lower degree than resident native species, and 2) AMF colonisation of native plants is lower in a community inhabited by an invasive species than in an uninvaded resident community. The two tests were run in semiarid temperate grasslands on grass (Poaceae) species, and the frequency and intensity of mycorrhizal colonisation, and the proportion of arbuscules and vesicles in plant roots have been measured. In the first test, grasses representing three classes of invasiveness were included: invasive species, resident species becoming abundant upon disturbance, and non-invasive native species. Each class contained one C3 and one C4 species. The AMF colonisation of the invasive Calamagrostis epigejos and Cynodon dactylon was consistently lower than that of the non-invasive native Chrysopogon gryllus and Bromus inermis, and contained fewer arbuscules than the post-disturbance domi- nant resident grasses Bothriochloa ischaemum and Brachypodium pinnatum. The C3 and C4 grasses behaved alike despite their displaced phenologies in these habitats. The second test compared AMF colonisation for sand grassland dominant grasses Festuca vaginata and Stipa borysthenica in stands invaded by either C. epigejos or C. dactylon, and in the uninvaded natural community. Resident grasses showed lower degree of AMF colonisation in the invaded stand compared to the uninvaded natural community with F. vaginata responding so to both invaders, while S. borysthenica responding to C. dactylon only. These results indicate that invasive grasses supposedly less reliant on AMF symbionts have the capacity of altering the soil mycorrhizal community in such a way that resident native species can establish a considerably reduced extent of the beneficial AMF associations, hence their growth, reproduction and ultimately abundance may decline. Accumulating evidence suggests that such indirect influences of invasive alien plants on resident native species mediated by AMF or other members of the soil biota is probably more the rule than the exception. Nomenclature: Tutin et al. (1964-1993). Abbreviations: AMF – Arbuscular Mycorrhizal Fungi; a% – percentage arbuscule occurrence of the AMF colonised root section; A% – percentage arbuscule occurrence of the whole root; F% – frequency of root segments colonised by AMF; M% – intensity of mycorrhizal colonisation; v% – percentage vesicle occurrence of the AMF colonised root section; V% – percent- age vesicle occurrence of the whole root. Introduction attention has turned towards the soil biota as a possible me- diator of plant-plant interactions in plant invasions (Klirono- Several, mutually non-exclusive hypotheses have been mos 2002, Fitter 2005). Arbuscural mycorrhizal fungi proposed to explain the success of invasive plants in their (AMF) are particularly important in this respect due to their new range (reviewed by Hierro et al. 2005, Mitchell et al. widespread occurrence and important role in communities 2006): the enemy release hypothesis (Keane and Crawley and ecosystems (Brundrett 1991). 2002), the novel weapons hypothesis (Callaway and Ride- nour 2004, Callaway et al. 2008, Wolfe et al. 2008), the evo- The majority of herbaceous plant species have AMF lution of increased competitive ability (EICA, Blossey and symbionts (Wang and Qiu 2006), yet data are insufficient Nötzold 1995, Bossdorf et al. 2005), the theory of fluctuating whether invasive species are associated with AMF (Pringle resource availability (Davis et al. 2000), the framework of et al. 2009). Although invasive plants may benefit from mu- vacant niches (Moles et al. 2008), and the facilitation of fur- tualistic relationships such as AMF symbioses (Richardson ther invaders as described in the invasional meltdown hy- et al. 2000, Shah et al. 2009b), it has been hypothesized that pothesis (Simberloff and Von Holle 1999). More recently, exotic plants without obligate dependence on an AMF sym- 68 Endresz et al. biont have greater chance to become invasive in the new al. (2004) demonstrated net carbon transfer from F. idahoen- community compared to those with strong AMF associations sis to the invasive plant. Interestingly, AMF appeared not to (Pringle et al. 2009). The carbon cost of sharing assimilates play obvious role in the invasion of the same Centaurea spe- with mycorrhizal fungi can be quite high (Douds et al. 1988). cies on early successional sand dunes at the Great Lakes, Thus, lacking tight interspecific associates not only allows USA (Emery and Rudgers 2012). Yet in other cases, the easier spread and establishment of the species, but also spares AMF community played a role in repelling plant invasion the plant the considerable cost of the fungal symbiont. As- (Stampe and Daehler 2003, Barto et al. 2010). similates saved in this way can be spent on growth, hence These examples clearly show that the role of AMF in resulting in greater competitive ability. Indeed, AMF asso- plant invasion is complex and context dependent. The aim of ciations were found less frequent among invasive species our study is to complement existing evidence and thus con- than native species (Vogelsang et al. 2004, Pringle et al. tribute to a deeper understanding of the problem. We in- 2009). Furthermore, rapid evolution of the plant’s decreased cluded only grass (Poaceae) species in our study to minimise reliance on AMF mutualists during invasion has been docu- taxonomic variation, and because AMF symbiosis is com- mented (Seifert et al. 2009). Notwithstanding, opposite find- ings have also been reported. The invasive species possessed mon in this family (see e.g. Wang and Qiu 2006). Further- more intense AMF infection than species in the resident com- more, grasses are dominant in grasslands, thus their contribu- munity (Marler et al. 1999, Greipsson and DiTommaso tion to ecosystem functions is decisive. We are aware of only 2006), and there are several invasive plants heavily colonised few studies dealing with the role of AMF symbionts in inva- by AMF in their new habitats (e.g. Ambrosia artemisiifolia, sion by grasses (Goodwin 1992, Hawkes et al. 2006, van der Kovács and Bagi 2001, Kovács and Szigetvári 2002, Fu- Putten et al. 2007). Specifically, we tested the following two manal et al. 2006). hypotheses. Several mechanisms involving AMF have been proposed H1) Roots of invasive species differ in the rate of AMF col- and experimentally tested that despite the low host-specifity onisation from the roots of resident native species. of arbuscular mycorrhizal associations, AMF can take part in H2) Roots of resident native species show different rate of plant invasions (mutualist facilitation, degraded mutualism; AMF colonisation in habitats infected by invasive species Mitchell et al. 2006, Reinhart and Callaway 2006, Shah et al. compared to habitats free from the invasive plant. 2009b, Vogelsang and Bever 2009). In its new habitat, the invasive plant can alter the soil microbial community (Kour- To examine these hypotheses, we conducted two sepa- tev et al. 2002, Batten et al. 2006, Greipsson and DiTommaso rate field tests in semiarid temperate grasslands in Hungary. 2006). This can be achieved directly through exuded allelo- These grasslands are key components of the natural forest- chemicals decimating indigenous AMFs (Stinson et al. 2006, steppe vegetation of East-Central Europe, cover substantial Zhang et al. 2007). For example, such a mechanism has been areas and have been greatly modified by human land uses reported for the European Alliaria petiolata in North Ameri- over centuries. The spread of invasive species is one of the can forests (Stinson et al. 2006, Hale et al. 2011, but see Barto most serious threats to these grasslands today (Molnár et al. et al. 2010). Alternatively, the soil microbial community may 2012). Data on AMF associations of plants in this vegetation be changed indirectly via the mass effect of the invasive plant started to appear in the last decade only (Kovács and Bagi being non-mycorrhizal or having different AMF symbionts, 2001, Kovács and Szigetvári 2002, Endresz et al. 2005), and hence local AMFs gradually attenuate (Niu et al. 2007, van particularly little is known about alien plants in this respect. der Putten et al. 2007, Vogelsang and Bever 2009, Meinhardt The grasses chosen for this study are common characteristic and Gehring 2012, Sanon et al. 2012). Different AMF spe- components of these grasslands and play different role in cies may have different colonisation phenologies (see e.g. vegetation dynamics. Calamagrostis